ABIR Response: IAIS Consultation Risk-Based Global Insurance Capital Standard (ICS) December 2014
February 16th, 2015
ABIR RESPONSE: IAIS CONSULTATION RISK-BASED GLOBAL INSURANCE CAPITAL STANDARD (ICS) DECEMBER 2014
www.iaisweb.org; electronically filed
The Association of Bermuda Insurers and Reinsurers (ABIR) appreciates the opportunity to provide comments on the latest IAIS Global Insurance Capital Standard (ICS) Consultation Paper. ABIR represents 21 of Bermuda’s international insurers and reinsurers that protect consumers around the world. With headquarters and operations in Bermuda and with operating subsidiaries in the United States and Europe, these carriers do business in more than 150 countries. Hence, ABIR has a keen interest in the development of the ICS to be applied to Internationally Active Insurance Groups (IAIGs).
Timing of the ICS needs to be expanded
We consider the current timing as unrealistic and challenging. We believe that the ICS process is outpacing critically important changes to jurisdictional solvency regimes which have been developed globally and have not yet been tested. We believe that the ICS could benefit greatly from the experience of existing group capital regimes such as Bermuda and Switzerland and changes over the next three to four years with the implementation of Solvency II and the development of capital rules for U.S. insurers; Asian insurers and Latin American/Mexican insurers.
Need for Testing, Impact and Cost/Benefit Analysis
A considerable further period of testing and impact analysis is needed before an ICS could be established. Such testing should encompass pre- and post-implementation testing and should reflect on the real experience that will be gained, and lessons learned from operating under prudential regimes currently being developed and implemented. Assessments on the effects of the ICS should not be limited to a quantitative analysis on an insurer level. Appropriate transition and phase-in periods would are necessary to avoid “cliff effects” and other unintended consequences. Field testing and market analyses should consider explicitly the incremental costs of implementing the ICS.
We would define a principles-based framework as one that wholly or in substantial part relies on standards set at the national or regional level; this definition aligns with the view that the goals of the ICS (i.e. policyholder protection and financial stability) can be met through local requirements, such as Solvency II, the Swiss Solvency Test, the Bermuda BSCR and emerging approaches to group capital in the U.S. and other jurisdictions. This means that local requirements should be considered and recognized as the development and implementation of the ICS progresses in order to avoid duplicative standards at the local level and globally.
The Purpose of the ICS should not be to raise capital/Level Playing Field
We believe that the ICS should not be intended as a capital-raising exercise for IAIGs and G-SIIs as a whole. Nor should the purpose of the ICS be designed to raise capital across the board. The use of the ICS as a capital-raising exercise would also exacerbate the level playing field issues. A different capital standard for IAIGs and non-IAIGs competing in the same markets with similar products would impact and distort insurer incentives, product availability and product cost. These impacts could have a detrimental effect on policyholders and policyholder protection.
Question 1: Are these principles appropriate as the foundation for a global consolidated insurance capital standard? Are any enhancements or modifications needed to the ICS Principles?
Response: The proposed principles should be aligned with the objective of developing an ICS for IAIGs (and G-SIIs) that focuses on meeting policyholder protection. Whilst the goal of having globally comparable risk-based measure of capital adequacy and comparability of outcomes across jurisdictions is admirable; we note that group capital standards have evolved over time in various jurisdictions (regions) and are still evolving in others to meet the demands of the markets which they represent. The ICS needs to be a ‘minimum benchmark of capital’ which can then be compared to existing jurisdictional group standards and reviewed and considered as to whether or not the outcomes are comparable. The main focus of the development of the ICS regulatory tool should be to enhance the protection of policyholders and be a means for supervisors to coordinate approaches within a group supervisory college.
Principle 2 – The main objective of the ICS are protection of policyholders and to contribute to financial stability
As stated above ABIR agrees that the primary goal of the ICS should be policyholder protection and to that end the IAIS should consider how the ICS attributes to improving policyholder protection vis-à-vis existing national group capital requirements. The ICS should provide the basis for supervisors to understand and discuss the capital thresholds that an IAIG meets compared to its existing regulatory group capital requirements.
Principle 8 – The ICS strikes an appropriate balance between risk sensitivity and simplicity
ABIR agrees that there should be a balance between risk-sensitivity and simplicity, but more importantly the ICS should be able to work in practice and appropriately reflects the risks to which insurers are exposed. To that end, ABIR strongly recommends that the IAIS provide appropriate recognition of diversification and risk-mitigation instruments utilized by groups (including re-insurance) which are key to achieving the level of risk-sensitivity within the ICS framework. Diversification and risk mitigation are fundamental aspects of the insurance business and are also closely linked to ICS Principle 6 on promoting sound risk management by IAIGs and G-SIIs.
Question 2: What does comparability mean for the ICS from your perspective?
Response: Comparability for a risk measure should be the ability to factor in both quantitative and qualitative data to assess capital adequacy of IAIGs across jurisdictions. It should not mean having the same data points for all IAIGs but the same processes and risks which an IAIG can be individually assessed on. It would be beneficial to understand what comparability means for the IAIS. Given the various approaches to valuation, accounting, etc. we are of the opinion that local group capital regimes that are consistent with an ICS framework on an outcomes-based analysis should be recognized as a suitable implementation of the ICS. In order to achieve a degree of comparability, the ICS should allow for the use of various valuations, accounting methodologies that are then adjusted to achieve similar outcomes. ABIR would welcome an approach whereby local regimes that are consistent with the ICS framework on an outcomes-based analysis are recognised as a suitable implementation of the ICS framework. We would like to specifically highlight that an option to use internal models, if these are part of local jurisdictional approaches, would contribute to the comparability of outcomes, by ensuring that all IAIGs’ actual risk profiles are accurately captured. Such an ICS delivering comparability of outcomes would form a useful basis for College discussions and would enhance mutual understanding. A College could reasonably take confidence from the knowledge that all supervisors present were working on capital requirements based on the same principles, with a common appreciation of risk and the value of risk mitigation actions. This might lead to improved trust between supervisors and therefore in due course to increased supervisory co-operation. ABIR believes that this approach to comparability best serves to create supervisory knowledge and understanding of an IAIG.
Section 5.2 GAAP with adjustments approach to valuation
Question 16: For the purpose of determining the ICS capital requirement, what adjustments, if any should be made to which local jurisdictional GAAP financial statements?
Response: ABIR supports ongoing determinations based on the field testing that would give considerations to local jurisdictional GAAP financial statements. It is important that the field testing examine this further. To that end, will the IAIS be inviting additional participants in the field testing to ensure a broad cross-section representative of the various jurisdictions/regions?
Section 6.2 Categorization of capital into tiers
Question 19: Should qualifying capital resources be classified in more than one or more than two tiers of capital? How many? And, if different from above, what key criteria should be used to determine tiering?
ABIR recognizes that not all capital is of the same quality, and so tiering may be appropriate. However, we would identify that discussion on tiering and limiting is potentially premature and existing local jurisdictional rules already exist in many cases. For the purpose of the ICS stress test capital resources should be identified and consideration of limitations and tiering are best assessed based on the results of the first test. For the ICS it is not particularly clear at this stage why more than one tier of capital is required. If the goal of the capital requirements is policyholder protection then all capital resources considered subordinate to policyholders should be available to meet capital requirements. Any possible reduction in value to these resources in times of stress or winding up could be recognized as part of the capital requirement calculation.
Section 7 ICS capital requirement
Question 37: Should the ICS capital requirement be developed so that it can be implemented as a PCR? If not, why not?
Response: ABIR does not agree with calibrating the ICS at the PCR level. The ICS should be viewed as a baseline solvency level considering that a global quantitative impact study and the necessary in depth testing will be difficult to achieve and conclude in the roughly two year implementation time line as laid out by the IAIS. More importantly, by calibrating to a baseline solvency level, the ICS meets the policyholder protection test. The ICS would allow for a benchmarking of the global group capital which allows the well tested and idiosyncratic jurisdictional regulatory group risk based capital requirements to remain in place until such time as logic argues for them to move to an extensively tested international norm. The benchmarking leads to additional insight which contributes to the evolution of the standard over time. This approach which can include an element of a self-executing test for equivalence or at least a parameter based testing of local regulatory group capital requirements against the ICS can allow an ICS to be widely adopted meeting the goals of the FSB and giving the IAIS a remarkable achievement.
Section 7.2.1 Risk Measure
Question 42: Which risk measure – VaR, Tail-VaR or another – is most appropriate for ICS capital requirement purposes? Why?
Response: ABIR would support continued field testing of both so that results can be considered and reviewed to make such determinations.
Section 7.3 Risk Mitigation
Question 49: Do the proposed principles adequately address the concept of risk mitigation? If not, which principles should be changed and why? What additional principles should the IAIS consider and why? What unintended consequences do the proposed principles create?
Response: ABIR would wish to confirm the recognition of reinsurance as risk mitigation instrument as noted above.
10.2 Use of internal models
Question 159: Should the IAIS permit the use of partial internal models for calculating elements of the ICS capital requirement? If so, for which elements of the ICS capital requirement should partial models be allowed? What are the advantages and disadvantages?
Response: ABIR supports the use of partial internal models subject to local regulatory approvals. Internal models provide greater risk sensitivity depending on the business written and in particular can provide more accurate reflections of the nature of the risks assumed.
Question 160: (Related to Q 161) should the IAIS permit the use of a full internal model for calculating the ICS capital requirement? What are the advantages and disadvantages?
Response: ABIR supports the use of Internal Models (partial or full) to determine solvency if these fit with local jurisdictional methods. The option to use internal models is very important for the ICS to avoid becoming hugely complex, while ensuring the ICS reflects the real risks across all the companies applying the ICS, enhancing the comparability of outcomes. Internal models provide insurance companies and supervisors with better insights into the firm’s distinctive risks and therefore promote sound risk management, in line with ICS Principle 6. The use of internal models could be subject to governance mechanisms and supervisory approval, as proposed in the consultation document.
We support an ICS that would permit the use of full or partial internal models. In our view, internal models facilitate a risk-sensitive approach to supervisory and insurers’ internal assessments of capital adequacy by considering an insurer’s idiosyncratic risk profile. Both full and partial internal model usage should be possible, as some insurers elect a full modeling approach, whereas others find modelling particularly useful for certain business lines, such as catastrophe risk modelling.The use of internal models for the calculation of regulatory capital requirements creates a linkage between insurers’ risk management practices and prudential measures and facilitates comparability provided that the key risk drivers are calibrated to the same level across insurers by more directly relating regulatory capital levels to the risk profile of the insurer, rather than relying on rough standardized measures that may not correlate well to the key risks of an insurer and fail to reflect the multiple layers of some insurance risks.
Developing an ICS that would not allow for the use of internal models or creating a floor for internal models based on a standardized approach would create disincentives for the continued development, maintenance, validation and improvement of insurance internal models, to the detriment of sound risk management (ICS Principle 6). Disallowing the use of models or creating a standardized floor would disassociate the measure of regulatory capital from the risk profile of the insurer, reducing the risk sensitivity of the ICS (ICS Principle 4). The calculation of a standardized floor could differ markedly across insurers under the different valuation approaches in use across jurisdictions, thus reducing comparability and creating level playing field issues.
The use of internal models should not be benchmarked against a standard approach, if only because internal models are subject to intense levels of documentation, validation and testing whereas a “standard approach” usually gives limited insight as to why the approach is fair and relevant to measurement of the risks at hand. This limited insight makes the benchmarking exercise appear arbitrary as it is often impossible to “reconcile” to a statistic that is opaque to every party other than the body that produced the “standard” approach.
Question 169: In order to allow for the use of internal models, what are the criteria to be set in order to provide a framework consistent with the ICS principles?
Response: ICP 17 provides a framework under which internal models can be used. We are supportive of the criteria articulated under ICP 17 for the use of internal models for regulatory capital purposes. For example, ICP 17 requires prior supervisory approval for the insurer’s use of an internal model; requires the insurer to adopt risk modelling techniques and approaches appropriate to the nature, scale and complexity of its risks; requires the insurer to validate an internal model by subjecting it to three tests: “statistical quality test”, “calibration test” and “use test”; and the insurer is required to demonstrate that the model is appropriate for regulatory capital purposes.